On Monday, Hartford, Conn., became the first city in the country to turn over the management of all of its public schools to a private company. Clearly the Hartford Board of Education recognized that public education must move beyond the old nostrums -- such as the notion that pouring more money into education will produce better-educated students. What's needed is a new ethic that forces schools to really perform, not just to keep on making the same mistakes. But will the rest of the country catch on too?

It should. The current approach just isn't working. Between 1970 and 1990, real U.S. education expenditures increased 80% to $5,400 per pupil from $3,000 (in 1992 dollars). But performance on the National Assessment of Educational Progress has been essentially flat since 1970. There is no reason to believe any improvement in education has occurred.

Five years ago, 13 university economists formed the Panel on the Economics of Educational Reform to find out what's behind U.S. schools' dismal performance. Our unanimous conclusion, outlined in a report being released today by the Brookings Institution, is that, as currently organized, schools are terribly inefficient: They continue to spend money in ways that do not consistently lead to high achievement. This lack of vision is seen in the perennial quest to lower pupil-teacher ratios.

Between 1970 and 1990, American schools managed to reduce the average pupil-teacher ratio to 17.2 from 22.3. Yet extensive evidence has demonstrated that there is no relationship between class size and student performance. Similarly, while the percentage of teachers with master's degrees rose from less than a quarter to more than half between 1961 and 1986, and while schools regularly pay teachers with master's degrees higher salaries, there is no evidence that these degrees produce better teaching.

There are two fundamental problems with public schools.

First they have virtually no incentives to encourage better student performance -- or to be concerned about costs. A teacher who produces exceptionally large gains in student performance generally sees little difference in compensation or career advancement. A superintendent whose schools spend less while keeping student achievement constant is unlikely to be rewarded.

The second basic problem is that schools make almost no effort to evaluate the success or failure of programs. Virtually all evaluation, if it exists, comes before a program is instituted; once in existence, a program will tend to remain in place unless it is obviously broken.

Our panel believes the answer lies in performance incentives, not more regulation or more spending. Regulation seeks to impose a centralized command and control system that specifies not only the goals but how they are to be achieved. Performance incentives, on the other hand, specify what is to be achieved and leave it up to those who run the schools to figure out how.

Education is inherently a highly decentralized activity. It is almost impossible to create a single set of regulations to identify, hire and mobilize America's almost three million teachers. If there is a single, glaring lesson to be learned from past attempts at school reform it is that no single overarching approach can solve the problems of every school. Policy makers should decentralize school systems and set performance standards.

That said, schools have begun to experiment with a number of promising performance-incentive schemes. For instance, charter schools enable teachers to set up new, innovative institutions. School choice and vouchers give students and their parents an important voice in determining whether schools are good by allowing them to decide which to attend. Merit pay for teachers and principals provides another approach. And then there is the Hartford model: contracting educational services to private firms.

The basic problem with instituting these new approaches is that we have very little evidence about what works and what doesn't. While Hartford is taking the lead in outside contracting, there is little evidence to suggest how best to structure such contracts. As for vouchers, there is some information about a program that has been running in Milwaukee for four years, but considerable uncertainty remains about how to apply the Milwaukee model elsewhere. The same holds with other performance-incentive schemes.

This lead to our panel's final recommendation. We need to dramatically expand our information base by introducing new evaluation schemes. Evaluation has become controversial because most people equate it with standardized tests. But the public shouldn't judge a school based on standardized tests alone. Other factors -- including education in the home, economic background, innate abilities and the like -- must also be taken into account. Schools should be evaluated not just on test scores but also on what they contribute to education -- the value added. When properly instituted, a value-added system may find out that some schools with high average scores are really contributing very little to students' performance.

Note what's missing from our panel's recommendations: a plea for more money for schools. We believe that school reform can and should be accomplished while holding real resources constant. Sufficient resources are available -- they're just being misapplied. Schools simply have to learn to become more disciplined.

---

Mr. Hanushek is professor of economics and public policy at the University of Rochester. He was chairman of the Panel on the Economics of Educational Reform.

---

      Panel on the Economics of Educational Reform
Charles S. Benson, University of California, Berkley
Richard B. Freeman, Harvard University
Dean T. Jamison, U.C.L.A.
Henry M. Levin, Stanford
Rebecca A. Maynard, University of Pennsylvania
Richard J. Murnane, Harvard University
Steven G. Rivkin, Amherst
Richard H. Sabot, Williams
Lewis C. Solomon, Milkin Institute 
 for Job and Capital Formation
Anita A. Summers, University of Pennsylvania
Finis Welch, Texas A&M
Barbara L. Wolfe, University of Wisconsin, Madison